JP2020155096A - Autonomous mobile device, program, method for steering autonomous mobile device, and method for adjusting autonomous mobile device - Google Patents

Abstract

To provide an autonomous mobile device that allows stable steering control even when the loading weight or the way of loading freight is changed.SOLUTION: An autonomous mobile device conveys an object of conveyance along a target track, and comprises: steering means 111 that controls steering on the basis of a steering command Wtgt based on a follow-up error dP from the target track; and adjustment means that generates disturbance Noise that disturbs the follow-up to the target track, and adjusts a parameter Param of the steering means on the basis of a behavior after the generation of the disturbance.SELECTED DRAWING: Figure 6

Description

The present invention relates to an autonomous mobile device, a program, a steering method of the autonomous mobile device, and an adjustment method of the autonomous mobile device.

Conventionally, it has been widely studied in the world to automate the transportation work in a warehouse by using an automatic guided vehicle (AGV) which is an autonomous mobile device. Specifically, a configuration is already known in which a connection mechanism is provided on an automatic guided vehicle, an object to be transported (basket vehicle) is connected to the automatic guided vehicle, and the vehicle is transported in a warehouse.

Patent Document 1 (Patent No. 26768691) discloses a technique for stabilizing line tracking in an automatic guided vehicle which is an autonomous mobile device.

However, according to the technique disclosed in Patent Document 1, the steering control parameter is a fixed value and cannot cope with the steering characteristics that fluctuate depending on the load weight and loading method of the cargo, and the steering control is not possible depending on the loading state. There is concern that it will become stable and oscillate. That is, there is room for improvement from the viewpoint of performing stable steering control even if the loading weight and loading method of the cargo change.

The present invention has been made in view of the above, and an object of the present invention is to perform stable steering control even if the loading weight and loading method of the cargo change.

In order to solve the above-mentioned problems and achieve the object, the present invention controls steering based on a steering command based on a tracking error with the target trajectory in an autonomous moving device that transports the object to be transported along the target trajectory. It is characterized by comprising a steering means and an adjusting means for generating a disturbance that disturbs the follow-up to a target trajectory and adjusting parameters of the steering means based on the behavior after the occurrence of the disturbance.

According to the present invention, there is an effect that stable steering control can be performed even if the loading weight and loading method of the cargo change. As a result, it has the effect of making the line tracking in the autonomous mobile device more stable.

FIG. 1 is an explanatory diagram showing an automatic guided vehicle and a basket carriage in the transport system according to the first embodiment. FIG. 2 is a perspective view showing an example in which the ID panel is arranged on the basket carriage. FIG. 3 is an explanatory diagram showing an example of a distribution warehouse where a transportation system is expected to be applied. FIG. 4 is a block diagram showing an example of the hardware configuration of the controller of the automatic guided vehicle. FIG. 5 is a block diagram showing an example of a functional configuration exhibited by a controller of an automatic guided vehicle. FIG. 6 is a diagram showing a configuration of steering means. FIG. 7 is a diagram showing a feedback control loop for steering control. FIG. 8 is a diagram showing a desirable open-loop characteristic (gain only) in the frequency characteristic of the control loop of FIG. FIG. 9 is a flowchart schematically showing a flow of determination processing in the stability determination means. FIG. 10 is a diagram showing a configuration of the adjusting means. FIG. 11 is a diagram showing a closed loop characteristic in the frequency characteristic of the control loop of the steering control shown in FIG. 7. FIG. 12 is a diagram showing an example of phase difference detection. FIG. 13 is a graph showing changes in vehicle body steering characteristics due to changes in cargo loading. FIG. 14 is a diagram showing a closed loop characteristic in the frequency characteristic of the control loop of the steering control shown in FIG. 7. FIG. 15 is a diagram showing a configuration of an automatic guided vehicle according to the second embodiment. FIG. 16 is a diagram showing an example of a data table of the control gain Param. FIG. 17 is a diagram showing a configuration of an automatic guided vehicle according to a third embodiment. FIG. 18 is a diagram showing a vehicle weight detection method. FIG. 19 is a diagram showing an example of a data table of the control gain Param. FIG. 20 is a diagram showing an overall configuration of an automatic guided vehicle provided in the transport system of the fourth embodiment. FIG. 21 is a diagram showing a line shape in the adjustment section. FIG. 22 is a diagram showing an example of a camera image according to the fourth embodiment. FIG. 23 is a diagram showing a configuration of a steering control unit according to the fourth embodiment. FIG. 24 is a diagram showing a control loop of steering control according to the fourth embodiment. FIG. 25 is a diagram showing an open loop characteristic of steering control according to the fourth embodiment. FIG. 26 is a diagram showing a configuration of an adjusting unit according to the fourth embodiment. FIG. 27 is a diagram showing phase difference detection in the fourth embodiment. FIG. 28 is a diagram showing changes in the closed loop characteristics in the fourth embodiment. FIG. 29 is a diagram showing an example of assembly variation in the fourth embodiment. FIG. 30 is a diagram showing changes in the open loop characteristics in the fourth embodiment. FIG. 31 is a flowchart showing a flow of operation of the optimization process according to the fourth embodiment. FIG. 32 is a diagram showing an example of the result of the optimization process according to the fourth embodiment.

Hereinafter, embodiments of an autonomous mobile device, a program, a steering method of the autonomous mobile device, and an adjustment method of the autonomous mobile device will be described in detail with reference to the accompanying drawings.

(First Embodiment)
FIG. 1 is an explanatory diagram showing a self-propelled robot 1 and a basket carriage 2 provided in the transport system according to the first embodiment. In this first embodiment, an automatic guided vehicle (AGV) that automatically transports the basket trolley 2 to a desired transport destination by automatically connecting and towing a towed trolley such as the basket trolley 2 to be connected. : Automated Guided Vehicle) is an example of a transfer system in which a self-propelled robot 1 is applied to an autonomous mobile device. The self-propelled robot 1 is an autonomous moving device having a function of automatically connecting to a basket carriage 2 for loading a transported object. As a result, the self-propelled robot 1 can be towed by a simple moving device without having a structure capable of loading, so that a large number of objects loaded on the basket trolley 2 can be transported. it can.

As shown in FIG. 1, the self-propelled robot 1 includes a robot body 100 which is a device body, a magnetic sensor 3, a controller 4 which is a detection device, a power source (battery) 6, a power motor 7, a motor driver 8, and a first. The range sensor 9, the connecting device 10, the driving wheel 71, the driven wheel 72, and the like, which are the sensors of the above. The range sensor 9 recognizes the surrounding environment of the self-propelled robot 1.

In the transport system of the first embodiment, a magnetic tape as a guide tape is installed on the floor surface of the path in which the self-propelled robot 1 can travel, and the magnetic tape is detected by using the magnetic sensor 3 to detect the magnetic tape. It can be recognized that 1 is located on a travelable route. The guidance method for installing the tape on the floor surface is not limited to a configuration using a magnetic tape (magnetic type), but may be a configuration using an optical tape (optical type). When an optical tape is used, a reflection sensor, an image sensor, or the like can be used instead of the magnetic sensor 3.

Further, in the transport system of the first embodiment, autonomous driving that recognizes the self-position can be performed by collating the two-dimensional or three-dimensional map with the detection result of the range sensor 9. The range finder 9 is a scanning laser distance sensor (laser range finder (LRF)) that irradiates an object with a laser beam and measures the distance from the reflected light to the object. Hereinafter, the range sensor 9 may be referred to as LRF9.

A stereo camera, a depth camera, or the like can also be used as a sensor used for recognizing the self-position by collating the detection result with a two-dimensional or three-dimensional map.

In the self-propelled robot 1, the controller 4 controls the drive of the power motor 7 via the motor driver 8 based on the detection results of the magnetic sensor 3 and the range sensor 9, and the power motor 7 rotates and drives the drive wheels 71. As a result, the self-propelled robot 1 autonomously travels.

Two drive wheels 71 are provided on the left and right to support the robot body 100. The drive wheels 71 are driven by the power motor 7 at left and right independent drive speeds MtgtL and MtgtR to move the self-propelled robot 1 forward, backward, and turn.

The power motor 7 rotationally drives the drive wheels 71 in accordance with the turning speed command Wtgt and the translational speed command Ftgt (not shown). The average speed of the outer circumferences of the left and right drive wheels 71 corresponds to the translational speed of the self-propelled robot 1, and the speed difference between the outer circumferences of the left and right drive wheels is a value proportional to the turning speed of the self-propelled robot 1. The drive speeds MtgtL and MtgtR of the left and right drive wheels 71 are calculated based on the following equations (1) and (2). The conversion coefficient km2 is a coefficient determined in proportion to the arrangement distance between the left and right drive wheels 71.

The rotation of the self-propelled robot 1 is not limited to the configuration realized by the difference in rotational speed between the left and right sides of the drive wheels 71, and may be configured to rotate the wheels at a steering angle, for example.

As shown in FIG. 1, the basket carriage 2 is a bottom plate 22 that holds the basket portion 20, casters 23 arranged at the four corners of the square bottom plate 22, and identification members arranged on the side surfaces of the basket portion 20. It includes an ID panel 21.

An ID panel 21 on which a recognition marker is displayed is attached to the basket trolley 2 placed at a predetermined position. The marker is coded with identification number information (ID information) of the basket carriage 2, transport destination information such as the transport position, and transport priority information using a retroreflective tape 21b (see FIG. 2) which is a strip-shaped member. ing. The identification number information (ID information) of the basket trolley 2 can be recognized by referring to a table or the like.

A marker reading device is installed in the self-propelled robot 1. The marker reading device includes a range sensor 9 which is an ID recognition means and a decoding unit. In the first embodiment, the controller 4 has a function as a decoding unit. The controller 4 recognizes the marker code from the detection result of the range sensor 9. The decoding unit of the controller 4 decodes the code information of the recognized marker to obtain the recognition number information, the transport destination information, and the priority information of the basket trolley 2.

In the first embodiment, the retroreflective tape 21b is used as a marker installed on the basket carriage 2. The self-propelled robot 1 reads using a range finder 9 such as a laser range finder (LRF) that acquires a distance from the surrounding environment. The controller 4 calculates the position coordinates of the ID panel 21 from the distance information between the ID panel 21 whose position is recognized by the range sensor 9 and the range sensor 9. The controller 4 controls the drive of the power motor 7 using the calculated position coordinates of the ID panel 21, so that the self-propelled robot 1 is positioned at a predetermined position in front of the ID panel 21 on the basket carriage 2.

Next, the ID panel 21 will be described in detail.

Here, FIG. 2 is a perspective view showing an example in which the ID panel 21 is arranged on the basket carriage 2. As shown in FIG. 2, the ID panel 21 is arranged at a substantially central portion in front of the basket carriage 2. More specifically, the ID panel 21 is arranged at a position facing the range sensor 9 of the self-propelled robot 1 (see FIG. 1). The ID panel 21 is removable from the basket carriage 2 and is installed by an operator at a predetermined position such as a skeleton (vertical bar) at the center of the basket carriage 2. Since the angle of the ID panel 21 is synonymous with the angle of the basket carriage 2, it is installed so as to be parallel to the front portion of the basket carriage 2.

In order for the self-propelled robot 1 to connect the basket trolley 2, the self-propelled robot 1 needs to detect the distance and angle between the basket trolley 2 and the self-propelled robot 1 and travel toward the basket trolley 2. .. However, when the range sensor 9 recognizes the shape of the basket trolley 2, it is difficult to accurately detect the distance and angle from the basket trolley 2 because the shape to be recognized changes depending on the loading status of the basket trolley 2. .. Therefore, in the first embodiment, the ID panel 21 is attached to the basket carriage 2 and the ID panel 21 is detected by the range sensor 9 mounted on the self-propelled robot 1.

Here, the ID panel 21 which is an identification member will be technically described. The ID panel 21 is an identification member for detecting and identifying a detection target by a detection device using an electromagnetic wave such as a laser range finder (LRF). In order to detect by electromagnetic waves or the like, the detection surface detected by the electromagnetic waves or the like (for example, the surface of the ID panel 21) is geometrically divided into at least three regions in the first direction (here, the horizontal direction) and divided. In the plurality of regions, the reflectances of at least adjacent regions with respect to electromagnetic waves and the like are set to be different.

Then, the detection device detects and identifies the detection target by detecting a specific pattern (signal) by utilizing the difference in the intensity of the reflected signal when irradiated with an electromagnetic wave or the like.

The transport system of the first embodiment using the self-propelled robot 1 automates the work of transporting a transport target with casters such as a basket cart 2 in a distribution warehouse or the like. The transfer operation by the self-propelled robot 1 is divided into the following three operations (1) to (3).
(1) Search and connection of transportation targets in the temporary storage area (2) Travel in the travel area (3) Search for storage location and unloading in the storage area

FIG. 3 is an explanatory diagram showing an example of a distribution warehouse 1000 to which a transport system is assumed to be applied. FIG. 3 shows the floor surface of the distribution warehouse 1000 as viewed from the ceiling side as a plan view. The XY plane shown in FIG. 3 is a plane parallel to the floor surface, and the Z axis indicates the height direction. In the distribution warehouse 1000 shown in FIG. 3, the temporary storage area A1 in (1) above is assumed to be, for example, a place where packages after picking (collection work in the warehouse) and unloaded packages are arranged. The storage area A2 in (3) above is assumed to be an area in front of the truck parking position for each direction of the truck berth, or an area in front of the elevator when the truck is transferred to another floor by an elevator or the like. Further, the traveling area A3 of the above (2) is assumed to be a place indicating a round-trip route between the temporary storage area A1 and the storage area A2 by the arrow in FIG.

The self-propelled robot 1 moves on the main line by a guidance system based on line recognition that recognizes a line of magnetic tape installed on the floor surface by a sensor. Further, the area mark 52 next to the line is detected to determine the area. Further, the ID panel 21 includes information on the storage area A2 as a transport destination and information on the priority order.

As shown in FIG. 3, a magnetic tape for guiding the self-propelled robot 1 is provided in a line shape in the traveling area A3, and a traveling line 51 on which the self-propelled robot 1 travels is provided. Further, at the start position and the end position of the temporary storage area A1 and the storage area A2 in the travel area A3, the area marks 52 are arranged in the vicinity of the travel line 51 to recognize in which area the self-propelled robot 1 is located. You can do it.

In the program executed by the self-propelled robot 1 described later, the operation can be specified for each area. The self-propelled robot 1 performs a connection operation of the basket trolley 2 in the temporary storage area A1 and a garage entry operation of the basket trolley 2 in the storage area A2. After the transportation of the basket trolley 2 to the storage area A2 is completed, the self-propelled robot 1 self-propells itself in the temporary storage area A1 where the basket trolley 2 is placed in order to carry the next basket trolley 2. Move by running.

In the first embodiment, the traveling area A3 is provided with the traveling line 51 using the magnetic tape for guiding the self-propelled robot 1, but the present invention is not limited to this. For example, the traveling line 51 is not essential in the traveling area A3, and area marks may be installed at predetermined intervals. In this case, the self-propelled robot 1 determines its own position from the rotation speeds of the driving wheels 71 and the driven wheels 72 and travels between the area marks.

In the first embodiment, the temporary storage area A1 and the storage area A2 are located immediately next to the traveling line 51. The self-propelled robot 1 searches within the temporary storage area A1 and the storage area A2 while traveling on the traveling line 51. When the basket trolley 2 to be transported is found in the temporary storage area A1, the operation shifts to the connection operation from the traveling line 51 to the basket trolley 2. Further, the storage area A2 is also searched for an empty address from the traveling line 51, and the basket trolley 2 is put into the garage.

In addition, in the distribution warehouse 1000 shown in FIG. 3, a plurality of retroreflective tapes 53, which are reflective materials, are installed at positions opposite the storage area A2 across the traveling line 51 and away from the storage area A2. ing. The plurality of retroreflective tapes 53 are installed at positions where the range sensor 9 of the self-propelled robot 1 can detect them. The self-propelled robot 1 estimates its own position based on the installation information of the plurality of retroreflective tapes 53.

Next, the controller 4 of the self-propelled robot 1 will be described.

Here, FIG. 4 is a block diagram showing an example of the hardware configuration of the controller 4 of the self-propelled robot 1. As shown in FIG. 4, the controller 4 includes a control device 11 such as a CPU (Central Processing Unit) and a GPU (Graphics Processing Unit), and a main storage device 12 such as a ROM (Read Only Memory) and a RAM (Random Access Memory). An auxiliary storage device 13 such as an SSD (Solid State Drive), a display device 14 such as a display, an input device 15 such as a keyboard, and a communication device 16 such as a wireless communication interface are provided. It has a hardware configuration that uses a computer.

The control device 11 controls the operation of the entire controller 4 (self-propelled robot 1) by executing various programs stored in the main storage device 12 and the auxiliary storage device 13, and realizes various functional units described later. ..

The program executed by the controller 4 of the self-propelled robot 1 is a file in an installable format or an executable format on a computer such as a CD-ROM, a flexible disk (FD), a CD-R, or a DVD (Digital Versatile Disk). It may be configured to be recorded and provided on a readable recording medium.

Further, the program executed by the controller 4 of the self-propelled robot 1 may be stored on a computer connected to a network such as the Internet and provided by downloading via the network. Further, the program executed by the controller 4 of the self-propelled robot 1 may be provided or distributed via a network such as the Internet.

Next, the function exhibited by the controller 4 of the self-propelled robot 1 will be described when the control device 11 of the controller 4 of the self-propelled robot 1 executes a program stored in the main storage device 12 and the auxiliary storage device 13. Here, the description of the conventionally known functions will be omitted, and the characteristic functions exhibited by the controller 4 of the self-propelled robot 1 of the first embodiment will be described in detail.

A part or all of the functions exhibited by the controller 4 of the self-propelled robot 1 may be configured by using a dedicated processing circuit such as an IC (Integrated Circuit).

FIG. 5 is a block diagram showing an example of a functional configuration exhibited by the controller 4 of the self-propelled robot 1. As shown in FIG. 5, the controller 4 of the self-propelled robot 1 includes steering means 111, stability determination means 112, and adjustment means 113.

The steering means 111 controls steering based on a steering command based on a tracking error with the target trajectory.

The stability determination means 112 determines the stability of the steering control of the steering means 111.

When the stability determining means 112 determines that the stability is insufficient, the adjusting means 113 adjusts the parameters of the steering means 111.

Next, the feedback control (steering control) of the steering of the self-propelled robot 1 will be described.

Generally, in the self-propelled robot 1 which is a line trace type automatic guided vehicle (AGV), there are wheel diameter errors and assembly errors of the left and right drive wheels, and especially in the case of the differential drive type. Since there are individual differences in the number of rotations of the motor, it may turn slightly even when the vehicle is instructed to go straight, so it is necessary to always steer left and right to correct the direction of travel even when going straight. Therefore, the self-propelled robot 1 performs feedback control for correcting the traveling direction by constantly steering left and right when tracing and moving on the orbit.

However, the steering characteristics of the self-propelled robot 1 that is connected to the object to be transported (basket 2) and autonomously moves varies depending on the load weight and loading method of the load loaded on the object to be transported (basket 2).

In the steering feedback control, the reciprocal of the steering cycle to the left and right (speed to steer to the left and right) is the frequency, and the coefficient that determines the magnitude of the steering angle to steer to the left and right according to the error with the trajectory is the gain. Become.

Therefore, it is important to perform steering control according to the load weight and loading method of the cargo.

First, the steering means 111 will be described.

Here, FIG. 6 is a diagram showing the configuration of the steering means 111. As shown in FIG. 6, the steering means 111 detects the line position error dP between the self-propelled robot 1 and the magnetic tape on the floor surface by using the magnetic sensor 3. The steering means 111 outputs a turning speed command Wtgt calculated by proportional differential control (PD control) based on the detected line position error dP. The error detection with the magnetic tape on the floor surface is not limited to the line position error dP, and may be configured to output an angle error or only an angle error.

Further, as shown in FIG. 6, the steering means 111 includes a disturbance adding unit 35, a proportional gain 31, a differential gain 32, a differential processing unit 33, and an output adding unit 34.

The disturbance addition unit 35 adds and outputs the line position error dP and the disturbance noise. Details of the disturbance noise will be described later.

The proportional gain 31 is output by multiplying the output of the disturbance addition unit 35 by the proportional parameter Kp. The proportional parameter Kp is changed according to the control gain Param described later. The control gain Param is a value of the proportional parameter Kp.

The differential gain 32 is output by multiplying the output of the proportional gain 31 by the differential parameter Kd. In the first embodiment, the differential parameter Kd is fixed, but may be changed by the control gain Param as in the proportional gain 31.

The differential processing unit 33 differentiates and outputs the output of the differential gain 32. The differential processing unit 33 may output the difference for each predetermined cycle.

The output addition unit 34 adds the output of the proportional gain 31 and the output of the differential processing unit 33, and outputs it as a turning speed command Wtgt.

Here, the proportional parameter Kp and the differential parameter Kd will be described with reference to FIGS. 7 and 8.

FIG. 7 is a diagram showing a feedback control loop for steering control. In FIG. 7, the steering control characteristic indicates the frequency characteristic of the steering means 111, and the vehicle body steering characteristic indicates the frequency characteristic from the turning speed command Wtgt to the actual position Pre. If the target follow-up error to the line is zero, the target position Ptgt is zero, and the sign-inverted value of the actual position Pre is synonymous with the line position error dP. The vehicle body steering characteristics shall be obtained in advance by actual measurement or simulation.

FIG. 8 is a diagram showing a desirable open-loop characteristic (gain only) in the frequency characteristic of the control loop of FIG. As shown in FIG. 8, the sum of the vehicle body steering characteristic (dotted line) and the steering control characteristic (dashed line) is the open loop characteristic (solid line). The proportional parameter Kp and the differential parameter Kd of the steering means 111 ensure sufficient stability so that the open loop characteristic becomes a characteristic expressed by a predetermined required specification value (crossing frequency fc, break point fl). Will be decided.

The configuration of the steering means 111 is not limited to PD control, and may be a feedback control system controller that updates the steering command at any time according to the tracking error to the target trajectory.

Next, the stability determination means 112 will be described.

The stability determination means 112 determines whether the stability of the steering control is sufficient (good) or insufficient (no) based on the line position error dP, and outputs the determination result Stbl.

Sufficient (good) or insufficient (no) stability of the steering control is determined by utilizing the fact that as the phase margin, which is an index of the stability of the feedback control, becomes insufficient, it becomes easier to oscillate at a constant cycle. However, the oscillating frequency is near the actual crossover frequency of the control loop (not necessarily equal to the crossover frequency fc of the design value). Therefore, in the first embodiment, the line position error dP is monitored and the determination process described later is periodically executed.

The stability determination is not limited to that described in the first embodiment, and may be configured to detect the magnitude of the amplitude by using, for example, a spectrum analysis by Fourier transform.

Here, FIG. 9 is a flowchart schematically showing the flow of the determination process in the stability determination means 112. As shown in FIG. 9, the stability determination means 112 first performs an initialization process (step S1). In the initialization process, for example, the determination result Stbl is set to "OK" when the stability of steering control is sufficient (good), and the continuous oscillation count is cleared (N = 0).

Next, the stability determining means 112 has a sign inversion of the line position error dP within a predetermined time (Yes in step S2), and the absolute value (amplitude) of the line position error dP is larger than the predetermined value (in step S3). Yes), the continuous oscillation count N is incremented by 1 (step S4). That is, the stability determination means 112 detects the periodic runout depending on the presence or absence of code inversion within a predetermined time. The predetermined time is determined based on the estimated value of the oscillating frequency.

The reason why the absolute value (amplitude) of the line position error dP is larger than the predetermined value (step S3) is added to the determination is to exclude the disturbance caused by the disturbance from the environment.

On the other hand, in the stability determination means 112, when there is no sign inversion of the line position error dP within a predetermined time (No in step S2) and a predetermined time has not elapsed since the previous sign inversion (No in step S5), step S2 Return to. Further, when the stability determination means 112 does not reverse the sign of the line position error dP within a predetermined time (No in step S2) and a predetermined time has elapsed from the previous sign inversion (Yes in step S5), step S1 Return to.

Further, in the stability determination means 112, the absolute value of the line position error dP is not larger than the predetermined value even if the sign of the line position error dP is reversed within a predetermined time (Yes in step S2) (in step S3). No), if a predetermined time has not passed since the previous code inversion (No in step S6), the process returns to step S3. Further, the stability determination means 112 has a sign inversion of the line position error dP within a predetermined time (Yes in step S2), and the absolute value of the line position error dP is not larger than the predetermined value (No in step S3). If a predetermined time has elapsed since the previous code inversion (Yes in step S6), the process returns to step S1.

The stability determination means 112 repeats the processes of steps S2 to S6 as described above until it is determined that the continuous oscillation count N is larger than the predetermined value (Yes in step S7).

When the stability determination means 112 determines that the continuous oscillation count N is larger than the predetermined value (Yes in step S7), the runout is continuous (that is, it is determined to be oscillation), and the stability of the steering control is insufficient (no). Assuming that, the determination result Stbl is set to “NG” (step S8) in which the stability of the steering control is insufficient (no), and the process is terminated.

Next, the adjusting means 113 will be described.

The adjusting means 113 automatically adjusts the parameters of the steering means 111 according to the determination result Stbl by the stability determining means 112. The automatic adjustment generates a disturbance noise to be injected into the control loop of the steering control, disturbs the line tracking of the self-propelled robot 1, and outputs a control gain Param that changes the parameter of the steering means 111 according to the behavior.

Here, FIG. 10 is a diagram showing the configuration of the adjusting means 113. As shown in FIG. 10, the adjusting means 113 includes a disturbance generating unit 61, an optimization processing unit 62, and a phase difference detecting unit 63.

The disturbance generation unit 61 outputs the disturbance noise in accordance with the disturbance generation request of the adjustment command Adj described later. The disturbance noise is a sine wave having a predetermined frequency fn. The disturbance noise is added to the control loop of the steering control in the steering means 111 and disturbs the follow-up to the line, but if the sine wave amplitude is reduced, there is no effect on practical running. Further, the disturbance noise is not limited to a sine wave, but may be a rectangular wave or a stepped signal having no periodicity.

If the determination result Stbl by the stability determination means 112 is negative, the optimization processing unit 62 outputs the adjustment command Adj, which is a command for generating disturbance noise, and outputs the control gain Param based on the phase difference dTH described later.

Here, the operation of optimizing the control gain Param will be described. FIG. 11 is a diagram showing a closed loop characteristic in the frequency characteristic of the control loop of the steering control shown in FIG. 7. In FIG. 11, the solid line indicates the target characteristic, the dotted line indicates the case where the vehicle body steering characteristic fluctuates on the gain decrease side, and the alternate long and short dash line indicates the case where the vehicle body steering characteristic fluctuates on the gain increase side. Is shown.

As shown in FIG. 11, when a disturbance noise of the noise frequency fn set near the crossover frequency fc is injected, the phase difference of the line position error dP corresponding to the response is the target phase difference Ph0 if there is no gain fluctuation. It can be seen from the lower plot of FIG. 11 that it advances from the target phase Ph0 as the gain increases and lags behind the target phase Ph0 as the gain decreases. The optimization process takes advantage of this property.

When the phase difference dTH is larger than the target phase Ph0, the optimization processing unit 62 lowers the value of the control gain Param. Further, the optimization processing unit 62 raises the value of the control gain Param when the phase difference dTH is smaller than the target phase Ph0. The optimization processing unit 62 repeats this operation until the phase difference dTH falls within a predetermined range centered on the target phase Ph0. As a result, it is possible to obtain a closed loop characteristic equivalent to that in the case of no gain fluctuation in FIG.

The phase difference detection unit 63 detects the phase difference between the line position error dP and the disturbance noise, and outputs the phase difference dTH.

Here, the phase difference detection in the phase difference detection unit 63 will be described. FIG. 12 is a diagram showing an example of phase difference detection. As shown in FIG. 12, the phase difference detection unit 63 measures the time difference Delay of the signal in which the line position error dP is sign-inverted and the zero cross point of the disturbance noise. In the first embodiment, since the target position Ptgt = 0, the sign inversion of the line position error dP corresponds to the actual position Pre. At this time, the line position error dP and the disturbance noise are sinusoidal waves having a frequency of fn, and the time difference Delay can be obtained with higher accuracy if it is the average value of several measurements.

Next, the phase difference detection unit 63 calculates and outputs the phase difference dTH from the time difference Delay by the following equation (3).

The phase difference detection by the phase difference detection unit 63 is not limited to the above-mentioned example. For example, a Fourier transform of only the Noise frequency fn, specifically, a line with a disturbance noise and a signal having an advanced quarter period phase. A method of multiplying the position error dP and integrating may also be used.

Here, the effect of the feedback control (steering control) of the steering of the self-propelled robot 1 in the first embodiment will be described.

Cargo is loaded on the object to be transported (basket cart 2) towed by the self-propelled robot 1. Depending on the load weight and loading position of the cargo, the vehicle weight, inertia (rotating system moment of inertia), and center of gravity (CG: Center of Gravity) position around the center of rotation will fluctuate compared to the single object to be transported (basket 2). .. As a result, the vehicle body steering characteristics (dotted line) shown in FIG. 8 fluctuate.

Here, FIG. 13 is a graph showing changes in vehicle body steering characteristics due to changes in cargo loading. As shown in FIG. 13, when the cargo load changes, that is, when the distance Lw between the center of gravity (CG) and the center of rotation (CR) of the object to be transported (basket 2) changes, the gain fluctuation of the vehicle body steering characteristics changes. Will occur, and the characteristics of the control loop will also fluctuate.

Here, the relationship between the gain characteristic fluctuation of the control loop and the stability will be described. FIG. 14 is a diagram showing a closed loop characteristic in the frequency characteristic of the control loop of the steering control shown in FIG. 7. In FIG. 14, the solid line indicates the target characteristic, the dotted line indicates the case where the vehicle body steering characteristic fluctuates on the gain decrease side, and the alternate long and short dash line indicates the case where the vehicle body steering characteristic fluctuates on the gain increase side. Is shown. The frequency at which the gain intersects 0 [dB] is defined as the crossover frequency fc, and the difference between the phase of the crossover frequency fc and −180 [deg] is defined as the phase margin PM. The phase margin PM is a measure of control stability.

As shown in FIG. 14, if there is a gain fluctuation in the vehicle body steering characteristics, the phase margin PM may decrease on both the increase / decrease side and the stability may decrease. When the phase margin PM decreases, the control becomes unstable and problems such as oscillation occur.

Therefore, in the first embodiment, by adjusting the control parameters of the steering means 111 by the adjusting means 113, the control loop is automatically adjusted to the target characteristics even if the gain fluctuates due to the change in cargo loading. Therefore, the self-propelled robot 1 can carry the object to be transported (the basket carriage 2).

(Effect of the first embodiment)
As described above, according to the first embodiment, the steering control is controlled by detecting the lack of stability of the steering control due to the load weight and the loading method of the transport object (cage trolley 2) towed by the self-propelled robot 1. By adding disturbance noise (signal) to the control loop of, the line tracking of the self-propelled robot 1 is intentionally disturbed to the extent that it does not affect the actual driving, and the steering control parameters are automatically adjusted according to the disturbed behavior. Because of the configuration, stable transportation is possible even if the load weight / loading method of the cargo changes, and the stability of steering control during running of the self-propelled robot 1 that transports the object to be transported (basket 2) is insufficient. It is possible to suppress the oscillation of steering feedback control (steering control), in other words, the wobbling of the self-propelled robot 1 during traveling.

Further, for example, when traveling in a narrow aisle in a warehouse, the influence of this wobbling becomes large, so that by suppressing the wobbling, the object to be transported (basket trolley 2) can be stably transported even in the narrow aisle. ..

Further, since the steering control parameters are automatically adjusted according to the disturbed behavior of the self-propelled robot 1 by adding disturbance noise (signal) to the steering control control loop, the weight of the loaded cargo and the like can be adjusted in advance. The above effects can be provided without information.

Further, since the disturbance noise (signal) applied to the control loop of the steering control is used as a periodic signal, the automatic adjustment of the steering control parameters can be accurately realized by averaging.

It should be noted that the automatic adjustment of the steering control parameters can be applied not only to the line trace type autonomous driving but also to the autonomous driving in which the self-propelled robot 1 itself determines the traveling line and travels.

Further, in the first embodiment, the cargo is loaded on the object to be transported (basket cart 2) towed by the self-propelled robot 1, but the present invention is not limited to this, and the cargo is loaded on the self-propelled robot 1. It may be directly loaded.

(Second Embodiment)
Next, a second embodiment will be described.

The transport system of the second embodiment detects information related to the weight of the cargo loaded on the self-propelled robot 1 and selects it based on the detected information of steering control parameters optimally obtained in advance. The point of automatic adjustment is different from that of the first embodiment. Hereinafter, in the description of the second embodiment, the description of the same part as that of the first embodiment will be omitted, and the parts different from the first embodiment will be described.

Here, FIG. 15 is a diagram showing the configuration of the self-propelled robot 1 according to the second embodiment. The self-propelled robot 1 shown in FIG. 15 is in a form of directly loading cargo. As shown in FIG. 15, the self-propelled robot 1 further includes a loading unit 110 for directly loading cargo, a weight detecting unit (front) 70f, a weight detecting unit (rear) 70r, and a loading information generating means 114. In FIG. 15, the driven wheel 72 of the self-propelled robot 1 is located in front of the traveling direction, and the driving wheel 71 is located behind in the traveling direction.

The weight detection unit (front) 70f is arranged below the loading unit 11 and as far forward as possible with respect to the traveling direction of the self-propelled robot 1, detects the load by the loading unit 11, and outputs it as the loading weight (front) mf. To do.

The weight detection unit (rear) 70r is arranged below the loading unit 11 and as far back as possible with respect to the traveling direction of the self-propelled robot 1, detects the load by the loading unit 11, and outputs it as the loaded weight (rear) mr. To do.

The load information generating means 114 generates and outputs the load information Info based on the load weight (front) mf and the load weight (rear) mr. The loading information Info shall include information on the weight and the position of the center of gravity of the cargo loaded on the loading unit 11. The weight of the cargo can be easily calculated from the sum of the load weight (front) mf and the load weight (rear) mr. The position of the center of gravity of the cargo can be calculated from the ratio of the loaded weight (front) mf and the loaded weight (rear) mr and the interval at which the weight detection unit (front) 70f and the weight detection unit (rear) 70r are arranged.

The adjusting means 113 outputs a control gain Param that changes the steering control parameters of the steering means 111 based on the loading information Info.

FIG. 16 is a diagram showing an example of a data table of the control gain Param. The adjusting unit 66 selects the value of the control gain Param according to the data table shown in FIG. The data table shown in FIG. 16 is a value optimally set for each combination of the cargo weight and the center of gravity position value included in the loading information Info, and is known information on the vehicle weight and the center of gravity position of the self-propelled robot 1. Including, it is set in advance by simulation and actual measurement.

(Effect of the second embodiment)
As described above, according to the second embodiment, the information related to the weight of the cargo loaded on the self-propelled robot 1 is detected, and the steering control parameters are optimized in advance based on the value of the detected information. Since it is configured to be selected from the obtained data table and automatically adjusted, stable transportation can be performed even if the loading weight / loading method of the cargo changes.

Further, since the optimum steering control parameters can be selected with high accuracy based on at least the information including the weight and the position of the center of gravity of the cargo, stable transportation of the cargo can be realized with high reliability.

In addition, since the weight of the loaded cargo is detected at multiple locations and the weight and center of gravity of the cargo are calculated, changes in the loading can be detected immediately, so stable transportation of the cargo can be achieved immediately even if the loading status changes. can do.

(Third Embodiment)
Next, a third embodiment will be described.

The transport system of the third embodiment detects information related to the total vehicle weight of the self-propelled robot 1 including the weight of the cargo loaded on the self-propelled robot 1, and performs steering control optimally obtained in advance. It differs from the first embodiment or the second embodiment in that the parameters are selected and automatically adjusted based on the detected information. Hereinafter, in the description of the third embodiment, the description of the same part as that of the first embodiment or the second embodiment is omitted, and the description is different from the first embodiment or the second embodiment. The part will be described.

Here, FIG. 17 is a diagram showing the configuration of the self-propelled robot 1 according to the third embodiment. The self-propelled robot 1 shown in FIG. 17 is in a form of directly loading cargo. As shown in FIG. 17, the self-propelled robot 1 further includes a communication means 120. In FIG. 17, the driven wheel 72 of the self-propelled robot 1 is located in front of the traveling direction, and the driving wheel 71 is located behind in the traveling direction.

The communication means 120 receives the total weight (front) Mf and the total weight (rear) Mr, which will be described later, from the transmission unit 202, which will be described later, by wireless communication, and outputs the received information.

Here, the total weight (front) Mf and the total weight (rear) Mr will be described. For example, the vehicle weight detection unit 201 is provided on the target trajectory of the self-propelled robot 1. The vehicle weight detection unit 201 has a function of measuring the load from above and outputs the measured value. It is desirable that the vehicle weight detection unit 201 be installed immediately after the place where the cargo is loaded and unloaded on the self-propelled robot 1 on the target track.

Here, FIG. 18 is a diagram showing a vehicle weight detection method. As shown in FIG. 18A, the vehicle weight detection unit 201 outputs the load measured when the driven wheel 72 of the self-propelled robot 1 comes to the upper part as the total vehicle weight (front) Mf. Further, as shown in FIG. 18B, the vehicle weight detection unit 201 outputs the load measured when the drive wheel 71 of the self-propelled robot 1 comes to the upper part as the total vehicle weight (rear) Mr.

Further, a transmission unit 202 for transmitting information by wireless communication is connected to the vehicle weight detection unit 201. The information transmitted by the transmission unit 202 is the total weight (front) Mf and the total weight (rear) Mr. The communication method of the transmission unit 202 is not limited to wireless communication, and may be any means that can transmit to the self-propelled robot 1.

The adjusting means 113 outputs a control gain Param that changes the steering control parameters of the steering means 111 based on the total weight (front) Mf and the total weight (rear) Mr received by the communication means 120.

FIG. 19 is a diagram showing an example of a data table of the control gain Param. The adjusting unit 67 selects the value of the control gain Param according to the data table shown in FIG. The data table shown in FIG. 19 is a value optimally set for each combination of the total weight (front) Mf and the total weight (rear) Mr, and is set in advance from simulation or actual measurement.

(Effect of the third embodiment)
As described above, according to the third embodiment, information related to the total vehicle weight of the self-propelled robot 1 including the weight of the cargo loaded on the self-propelled robot 1 is detected, and based on the value of the detected information. Since the steering control parameters are selected from the data table optimally obtained in advance and automatically adjusted, the vehicle weight detection unit 201 can be shared by a plurality of self-propelled robots 1, so that the cargo can be inexpensively configured. Stable transportation can be realized.

(Fourth Embodiment)
Next, a fourth embodiment will be described.

In the case of the technique disclosed in the above-mentioned Patent Document 1 (Patent No. 26768691), the steering control parameter is a fixed value. For this reason, it becomes difficult to deal with the vehicle body steering characteristics or line tracking error detection characteristics that fluctuate depending on the load weight or loading method of the cargo and fluctuate due to variations in parts or assembly, and steering control becomes unstable and oscillates depending on the loading state. There is concern about doing so.

That is, in the case of the technique disclosed in Patent Document 1, stable steering control is possible even if the loading weight or loading method of the cargo changes, or even if there are variations in parts or assembly. There is room for improvement from this perspective.

In the case of the transport system of the fourth embodiment, stable steering control is possible even if the loading weight or loading method of the cargo changes, or even if there are variations in parts or assembly, and autonomous movement is possible. It enables more stable line tracking in the device.

An automatic guided vehicle provided in the transport system of the fourth embodiment and an adjustment method thereof will be described. This automatic guided vehicle loads cargo and follows a target track laid on the floor or the ground. The target trajectory is formed, for example, by magnetically or optically detectable lines or markers. As an example, in the case of the fourth embodiment, a line that can be photographed by a camera is used.

FIG. 20 is a diagram showing a configuration of a self-propelled robot 210 provided in the transfer system of the fourth embodiment. The self-propelled robot 210 is an automated guided vehicle (AGV) including a loading unit 211, a universal caster 212, a position detection unit 220, a steering control unit 230, a motor unit 241 and a drive wheel 242 and an adjustment unit 260. , An example of an autonomous mobile device.

FIG. 21 is a view of the line 225, which is the target trajectory followed by the self-propelled robot 210, as viewed from above. The line 225 is composed of substantially continuous lines, and includes an adjustment section having a periodic shape in a part thereof. The periodic shape is one of a sine wave shown in shape A, a zigzag shape shown in shape B, and a shape in which a plurality of straight lines shown in shape C are arranged discontinuously and alternately (staggered arrangement). As long as the shape has periodicity, shapes other than shapes A to C may be used.

The swing width Amp of the periodic shape is a size that falls within the detectable range of the position detection unit 220 described later. Although it is an example, in the fourth embodiment, the size is within the field of view of the camera described later.

The length Lall of the entire adjustment section and the cycle length Lprd, which is the length per cycle, will be described later. The line 225 is an example of a target trajectory, and the shapes A, B, and C are examples of shapes having periodicity.

The adjustment mark 226 is arranged near the start point and the end point of the adjustment section. As an example, at the start point of the adjustment section, the adjustment mark 226 is arranged to the right of the line 225 with respect to the traveling direction of the self-propelled robot 210. Further, at the end point of the adjustment section, the adjustment mark 226 is arranged on the left side of the line 225 with respect to the traveling direction of the self-propelled robot 210. The adjustment mark 226 may have any shape and arrangement as long as the position detection unit 220 can detect the start point and the end point of the adjustment section.

Next, the configuration of the self-propelled robot 210 will be described. The loading unit 211 is a place for loading cargo in the self-propelled robot 210. A basket or enclosure is common, but the shape is not particularly limited. The universal caster 212 is a wheel and is rotatably provided so as to support the vehicle body of the self-propelled robot 210 and not to hinder the traveling and steering of the drive wheels 242 described later. The position detection unit 220 includes a camera, photographs the line 225 laid on the floor, and detects an error between the self-propelled robot 210 and the line 225 from the captured image.

FIG. 22 is a diagram showing an example of an image taken by the camera of the position detection unit 220. The top and bottom of FIG. 22 correspond to the front and back of the self-propelled robot 210. Further, the center in the horizontal direction is configured to coincide with the center of the vehicle body of the self-propelled robot 210. The position detection unit 220 calculates the distance between the center of the line image 221 of the captured line 225 and the center in the horizontal direction by image processing, and outputs it as a line position error dP.

The adjustment mark image 222 is a photographed adjustment mark 226. The position detection unit 220 detects the adjustment mark 226 by image processing. The position detection unit 220 is the adjustment mark 226 at the start point or the adjustment mark 226 at the end point, depending on whether the adjustment mark image 222 is on the left side or the right side of the line image 221. Is determined. The fact that the adjustment mark image 222 exists on the left side with respect to the line image 221 means that the adjustment mark image 222 is the adjustment mark image 222 at the start point. In this case, the position detection unit 220 outputs an adjustment start command (adjustment command Trg). On the other hand, the fact that the adjustment mark image 222 exists on the right side with respect to the line image 221 means that the adjustment mark image 222 is the adjustment mark image 222 at the end point. In this case, the position detection unit 220 outputs an adjustment end command (adjustment command Trg).

The error detection with the line is not limited to the line position error dP, and the angle of the line image 221 may be calculated and the angle error may be output, or only the angle error may be output.

The steering control unit 230 outputs a turning speed command Wtgt calculated by proportional differential control (PD (Proportional Differential) control) based on the line position error dP. The steering control unit 230 has the configuration shown in FIG. 23. In FIG. 23, the proportional gain 231 is output by multiplying the line position error dP by the proportional parameter Kp. The proportional parameter Kp is changed by the control gain Param described later. The control gain Param is a value of the proportional parameter Kp.

The differential gain 232 is output by multiplying the output of the proportional gain 231 by the differential parameter Kd. The differential parameter Kd may be fixed or variable by the control gain Param as in the proportional gain 31.

The differential processing unit 233 differentiates and outputs the output of the differential gain 232. The difference for each predetermined cycle may be output.

The output addition unit 234 adds the output of the proportional gain 231 and the output of the differential processing unit 233 and outputs the output as the turning speed designation Wtgt.

The proportional parameter Kp and the differential parameter Kd will be described with reference to FIGS. 24 and 25. FIG. 24 is a block diagram of a feedback control loop for steering control. The error detection characteristic 252 indicates the frequency characteristic of the position detection unit 220, and the steering control characteristic 253 indicates the frequency characteristic of the steering control unit 230. Further, the vehicle body steering characteristic 254 indicates a frequency characteristic from the turning speed command Wtgt to the actual position Pre. If the target follow-up error to the line is zero, the target position Ptgt is zero, and the actual position Pre is sign-inverted and supplied to the adder 251. Then, the value detected by the position detection unit 220 becomes the line position error dP.

FIG. 25 is a diagram showing desirable open-loop characteristics (gain only) in the frequency characteristics of the control loop of FIG. 24. The sum of the error detection characteristic (dashed line), the vehicle body steering characteristic (dotted line), and the steering control characteristic (dashed line) is the open loop characteristic (solid line). The open-loop characteristic becomes a characteristic expressed by a predetermined required specification value (crossing frequency fc, break point fl), and after ensuring sufficient stability, the proportional parameter Kp and the differential parameter Kd of the steering control unit are used. Is determined.

The error detection characteristic is calculated from the distance between the camera and the road surface of the position detection unit 220, the visual field range of the camera, and the like. In addition, the vehicle body steering characteristics can be obtained in advance by actual measurement or simulation. Further, the steering control unit 230 may be configured to perform feedback control for updating the steering command at any time according to, for example, a follow-up error to the target trajectory, in addition to the configuration for performing PD control.

Drive wheels 242 provided on the left and right sides of the self-propelled robot 210 support the vehicle body of the self-propelled robot 210. The drive wheels 242 are driven by a motor unit described later at left and right independent drive speeds MtgtL and MtgtR. As a result, forward control, backward control, and turning control of the self-propelled robot 210 are performed.

The motor unit 241 rotationally drives the drive wheels in accordance with the turning speed command Wtgt and the translation speed command Ftgt. The average speed of the outer circumferences of the left and right drive wheels 242 corresponds to the translational speed of the self-propelled robot 210, and the speed difference between the outer circumferences of the left and right drive wheels is a value proportional to the turning speed of the self-propelled robot 210. The left and right drive speeds MtgtL and MtgtR are calculated based on the following equations (4) and (5).

換算係数ｋｍ２は、左右の駆動輪４２間の配置距離に比例して決まる係数である。なお、モータユニット２４１及び駆動輪２４２は、操舵部の一例である。

The conversion coefficient km2 is a coefficient determined in proportion to the arrangement distance between the left and right drive wheels 42. The motor unit 241 and the drive wheels 242 are examples of the steering unit.

In addition to the configuration realized by the difference in rotational speed between the left and right drive wheels 42 of the present embodiment, the self-propelled robot 210 may be rotated by, for example, a steering angle.

The adjusting unit 260 automatically adjusts the parameters of the steering control unit 230 according to the adjustment command Trg. In the automatic adjustment, the shape A, shape B, and shape C of the line 225 intentionally disturb the target to be followed, and the control gain changes the parameters of the steering control unit 230 according to the behavior of the self-propelled robot 210 at that time. Output Param.

FIG. 26 shows the configuration of the adjusting unit 260. The adjusting unit 260 includes an optimization processing unit 262 and a phase difference detecting unit 263. The optimization processing unit 262 receives a command to start adjustment by the adjustment command Trg, and starts outputting a control gain Param based on the phase difference dTH described later. Further, the optimization processing unit 262 receives a command to end the adjustment by the adjustment command Trg, and stops updating the control gain Param.

The operation of the control gain Param optimization processing in the optimization processing unit 262 will be described with reference to FIG. 28. FIG. 28 shows a closed loop characteristic in the frequency characteristic of the control loop of the steering control shown in FIG. 24. The solid line in FIG. 28 shows the target frequency characteristic, and the dotted line shows the frequency characteristic when there is a fluctuation on the gain reduction side in the error detection characteristic or the vehicle body steering characteristic. In addition, the alternate long and short dash line shows the frequency characteristics when there is a fluctuation on the gain increase side.

When a disturbance of the frequency fn set near the crossover frequency fc is injected into the control loop, the phase difference from the line position error dP corresponding to the response is the target phase difference Ph0 if there is no gain fluctuation, whereas the gain As it increases, it advances from the target phase Ph0. On the contrary, the phase difference from the line position error dP becomes later than the target phase Ph0 as the gain decreases. The optimization processing of the control gain Param in the optimization processing unit 262 is performed by utilizing this characteristic.

That is, the optimization processing unit 262 lowers the value of the control gain Param when the input phase difference dTH is larger than the target phase Ph0. On the contrary, when the phase difference dTH is smaller than the target phase Ph0, the value of the control gain Param is increased. The optimization processing unit 262 repeatedly executes this operation until the phase difference dTH falls within a predetermined range centered on the target phase Ph0. As a result, it is possible to obtain a closed loop characteristic equivalent to that in the case of no gain fluctuation in FIG.

Next, the phase difference detection unit 263 detects the phase difference between the above-mentioned disturbance and the line position error dP, and outputs the phase difference dTH. First, injection of disturbance into the control loop will be described. A disturbance corresponding to the shape of the line 225 is injected into the control loop. This corresponds to the self-propelled robot 210 passing through the sections of the shape A, the shape B, and the shape C of the line 225 at a substantially constant speed. At this time, the disturbance frequency fn can be calculated by the following equation (6).

Therefore, the translation speed command Ftgt0 in the adjustment section is set to a predetermined value in advance so that the disturbance frequency fn is near the crossover frequency fc, and the predetermined period length Lprd of the shapes A, B, and C is set. Line 225 is formed so as to be.

Next, the reference waveform Ref of the disturbance used for calculating the phase difference dP is calculated. The elapsed time from receiving the adjustment start command by the adjustment command Trg is calculated by the following equation (7) as the time tadj from the adjustment start. Since only the phase is focused on, the value of the amplitude Aref of the reference waveform may be any positive positive value.

Next, the calculation of the phase difference dTH will be described. As shown in FIG. 27, the signal in which the line position error dP (thick line) is sign-inverted and the time difference Delay of the zero cross point of the reference waveform Ref are measured. In the fourth embodiment, since the target position Ptgt = 0, when the sign of the line position error dP is inverted, it corresponds to the actual position Pre. At this time, even if the line 225 has a non-sinusoidal shape of shape B or C, the follow-up waveform is blunted by the control loop, so that the line position error dP is substantially sinusoidal. By using the average value of several measurements as the time difference Delay, a highly accurate value can be used.

Next, the phase difference dTH is calculated from the time difference Delay by the following equation (8), and the value is output. The phase difference dTH shall keep the previous value until the next update.

The phase difference detection may be, for example, a Fourier transform process of only the disturbance frequency fn, which is a process of multiplying the line position error dP and integrating the reference waveform Ref and the signal with the advanced 1/4 period phase.

Further, in the case of the above-described embodiment, the cargo is loaded on the upper surface of the self-propelled robot 210, but the cargo may be loaded on the trolley towed by the self-propelled robot 210.

Next, a specific example of such an automatic adjustment operation will be described. First, various conditions are shown. The specifications of the control loop for the steering control of the self-propelled robot 210 are as follows.

Crossing frequency fc = 0.33 [Hz]
Break point fl = 0.1 [Hz]

In such a specification, when the disturbance frequency fn = 0.4 [Hz], the control gain Param and the target phase Ph0 are as follows.

Control gain Param = 1.88 (actual measurement)
Target phase Ph0 = -60 [deg] (actual measurement)

Further, the periodic shape of the line 225 of the adjustment section is the shape C shown in FIG. 21, the swing width Amp is 0.05 [m], and the translation speed command Ftgt0 in the adjustment section is 0.15 [m / s]. The number of periodic shapes was set to 12. In this case, the period length Lprd and the adjustment section length All are as follows.

Cycle length Lprd = Ftgt / fn = 0.375 [m]
Length of adjustment section All = Lprd × 12 = 4.5 [m]

There is no limit to the number of periodic shapes, but this time, the number of standby sections immediately after the adjustment section is set to 2, and the number of sections to be repeatedly adjusted is set to 10.

The optimization processing unit 262 automatically adjusts the parameters of the steering control unit 230 by periodically executing the operation shown in the flowchart of FIG. 31 based on such various conditions. The execution cycle is preferably sufficiently shorter than the disturbance cycle, and in this example, one disturbance cycle is set to 1/50 of "0.02 (1 / fn) [s]". Here, the adjustment command Trg is at a high level (Hi) from the command for starting adjustment to the command for ending adjustment, that is, while the self-propelled robot 210 is traveling in the adjustment section, and the adjustment is adjusted otherwise. The command Trg is the low level "Lo".

In step S11, the optimization processing unit 262 determines whether or not it is an adjustment section. The optimization processing unit 262 proceeds to step S12 when the adjustment command Trg is Hi (step S11: Yes), and proceeds to step S16 otherwise (step S11: No).

In step S12, the optimization processing unit 262 secures a waiting time from entering the adjustment section until the behavior of the self-propelled robot 210 settles down (steady-state waiting). This is because the behavior is disturbed immediately after the self-propelled robot 210 invades the periodic shape portion from the straight portion of the line 225, and the phase difference dTH is not helpful. The cycle count n for determining the steady wait indicates the number of periodic shapes passed by the self-propelled robot 210. The standby count value n0 is a threshold value indicating how many periodic shapes must be passed after entering the adjustment section to start phase difference detection, and as an example, n0 = 2.

When the optimization processing unit 262 determines that the behavior of the self-propelled robot 210 has settled down steadily (step S12: Yes), the process proceeds to step S13, and when the behavior of the self-propelled robot 210 does not settle down steadily (step S12). : No), the entire process of the flowchart of FIG. 31 is terminated as it is.

In step S13, the optimization processing unit 262 determines that one new periodic shape has passed, based on whether or not the phase difference dTH is updated. When the phase difference dTH is updated (step S13: Yes), the optimization processing unit 262 advances the processing to step S14, and otherwise (step S13: No) ends all the processing of the flowchart of FIG. 31. To do.

In step S14, the optimization processing unit 262 calculates and updates the control gain Param based on the phase difference dTH according to the following equation (9).

Although it is an example, the adjustment coefficient α = 0.025 when the convergence is roughly performed in about 10 times.

Finally, in step S15, since one new periodic shape has been passed, the optimization processing unit 262 increments the adjustment count n. In step S6, since it is outside the adjustment section, the optimization processing unit 262 clears the adjustment count n and ends all the processing of the flowchart of FIG. 31.

FIG. 32 shows the phase difference dTH and the phase difference dTH when the initial value of the control gain Param is set to 3.76 (1.88 × 2), which is twice the target value, and automatic adjustment is performed under the above-mentioned conditions. The transition of the control gain Param is shown. In FIG. 32, the upper plot shows the time tadj from the start of adjustment on the horizontal axis and the detected phase difference dTH on the vertical axis. It can be seen that the self-propelled robot 210 starts detecting the phase difference dTH after the waiting time after entering the adjustment section, and asymptotically approaches the target phase difference Ph0 by automatic adjustment.

In FIG. 32, in the lower plot, the horizontal axis represents the time tadj from the start of adjustment, and the vertical axis represents the control parameter Param. As described in the flowchart of FIG. 31, the value is updated every time the phase difference dTH is updated by the automatic adjustment process of the parameter of the steering control unit 230, and the value is asymptotically approached to the target value "1.88". It can be seen that the street automatic adjustment processing is being performed.

(Effect of the fourth embodiment)
Cargo is loaded on the self-propelled robot 210. The vehicle weight, inertia around the center of rotation, and the CG position of the center of gravity vary depending on the load weight and the load position as compared with the self-propelled robot 210 alone. As a result, the vehicle body steering characteristics (dotted line) shown in FIG. 25 fluctuate, and the control loop characteristics also fluctuate.

In the self-propelled robot 210, the error detection characteristic (two-dot chain line) and the vehicle body steering characteristic characteristic (dotted line) shown in FIG. 25 fluctuate due to variations in parts or assembly. As for the variation of the parts, for example, if the wheel diameter of the drive wheel 42 varies, the vehicle body steering characteristics fluctuate. As for the variation in assembly, for example, as shown in FIG. 29, the distance between the road surface and the camera changes due to an error in the mounting posture of the camera of the position detection unit 20 (left: ideal, right: with error), and error detection is performed. The characteristics fluctuate. In this way, since the error detection characteristic or the vehicle body steering characteristic changes, the characteristic of the control loop also changes.

The relationship between the characteristic variation and the stability in the control loop will be described with reference to FIG. FIG. 30 shows an open loop characteristic in the frequency characteristic of the control loop shown in FIG. 24, and similarly to FIG. 28, the solid line shows the target characteristic, and the dotted line shows the fluctuation of the vehicle body steering characteristic on the gain reduction side. The case is shown as an example in which the alternate long and short dash line has a fluctuation on the gain increasing side in the vehicle body steering characteristics. The frequency at which the gain intersects 0 [dB] is the crossover frequency fc, and the difference between the phase of the crossover frequency fc and −180 [deg] is the phase margin PM. This phase margin PM serves as a measure of control stability.

If there is a gain fluctuation as shown in FIG. 30, there is a possibility that the phase margin PM decreases on both the increase / decrease side and the stability decreases. If the phase margin PM decreases, control becomes unstable and problems such as oscillation may occur.

Therefore, as described in the fourth embodiment, the self-propelled robot 210 that moves along the target trajectory by steering control is based on the behavior during movement in a specific section of the target trajectory having a periodic shape. And adjust the control parameters. As a result, the control loop can be adjusted to the target characteristics even if the cargo load changes or the gain fluctuates due to variations in parts or assembly. Therefore, the line tracking of the self-propelled robot 210 can be made more stable, and stable transportation can be performed.

Finally, each of the above embodiments is presented as an example and is not intended to limit the scope of the invention. This novel embodiment can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the gist of the invention. In addition, each embodiment and modifications of each embodiment are included in the scope and gist of the invention, and are included in the scope of the invention described in the claims and the equivalent scope thereof.

1 Self-propelled robot 2 Transport object 111 Steering means 112 Stability judgment means 113 Adjustment means 120 Communication means 210 Self-propelled robot 220 Position detection unit 230 Steering control unit 241 Motor unit 260 Adjustment unit 262 Optimization processing unit

Japanese Patent No. 26768691

Claims (20)

In an autonomous moving device that transports an object to be transported along a target trajectory,
A steering means that controls steering based on a steering command based on a tracking error with the target trajectory, and
An autonomous moving device including an adjusting means for generating a disturbance that disturbs the follow-up to the target trajectory and adjusting parameters of the steering means based on the behavior after the occurrence of the disturbance. A stability determining means for determining the stability of the steering control of the steering means is provided.
The autonomous mobile device according to claim 1, wherein the adjusting means performs the adjustment when it is determined by the stability determining means that the stability is insufficient. The autonomous mobile device according to claim 2, wherein the adjusting means generates the disturbance having periodicity. In an autonomous moving device that transports an object to be transported along a target trajectory,
A steering means that controls steering based on a steering command based on a tracking error with the target trajectory, and
An autonomous moving device comprising: adjusting means for adjusting parameters of the steering means based on loading information indicating a loading state of the object to be conveyed. The autonomous moving device according to claim 4, wherein the loading information includes at least the weight of the object to be transported and the position of the center of gravity. The autonomous mobile device according to claim 4 or 5, further comprising a communication means for receiving the loading information from the outside. A computer that controls an autonomous moving device that transports an object to be transported along a target trajectory.
A steering means that controls steering based on a steering command based on a tracking error with the target trajectory, and
A program for generating a disturbance that disturbs the follow-up to the target trajectory and functioning as an adjusting means for adjusting the parameters of the steering means based on the behavior after the occurrence of the disturbance. A computer that controls an autonomous moving device that transports an object to be transported along a target trajectory.
A steering means that controls steering based on a steering command based on a tracking error with the target trajectory, and
A program for functioning as an adjusting means for adjusting parameters of the steering means based on loading information indicating a loading state of the transported object. It is a steering method of an autonomous moving device that transports an object to be transported along a target trajectory.
A steering process that controls steering based on a steering command based on a tracking error with the target trajectory, and
A steering method for an autonomous moving device, which comprises an adjustment step of generating a disturbance that disturbs the follow-up to the target trajectory and adjusting parameters of the steering step based on the behavior after the disturbance occurs. It is a steering method of an autonomous moving device that transports an object to be transported along a target trajectory.
A steering process that controls steering based on a steering command based on a tracking error with the target trajectory, and
A steering method for an autonomous moving device, comprising: an adjustment step of adjusting parameters of the steering step based on loading information indicating a loading state of the transported object. It is a method of adjusting an autonomous moving device that is equipped with a steering means that controls steering based on a tracking error with a target trajectory and moves along the target trajectory.
An autonomous moving device, characterized in that the adjusting unit adjusts parameters of the steering means based on the behavior of the autonomous moving device in a predetermined section in which the shape of the target trajectory is not a shape connecting two points at the shortest distance. Adjustment method. The method for adjusting an autonomous mobile device according to claim 11, wherein the shape of the target trajectory in the predetermined section has a periodicity with the line connecting the two points at the shortest distance as the center line. The method for adjusting an autonomous mobile device according to claim 12, wherein the shape having periodicity is a sine wave. The method for adjusting an autonomous mobile device according to claim 12, wherein the shape having periodicity is a zigzag shape. The method for adjusting an autonomous mobile device according to claim 12, wherein the shape having periodicity is a shape in which a plurality of straight lines are arranged discontinuously and alternately. It is an autonomous moving device that has steering means that controls steering based on the tracking error with the target trajectory and moves along the target trajectory.
An autonomous moving device having an adjusting unit that adjusts parameters of the steering means based on the behavior of the autonomous moving device in a predetermined section in which the shape of the target trajectory is not a shape connecting two points at the shortest distance. The autonomous moving device according to claim 16, wherein the shape of the target trajectory in the predetermined section has a periodicity with the line connecting the two points at the shortest distance as the center line. The autonomous mobile device according to claim 17, wherein the shape having periodicity is a sine wave. The autonomous mobile device according to claim 17, wherein the shape having periodicity is a zigzag shape. The autonomous moving device according to claim 17, wherein the shape having periodicity is a shape in which a plurality of straight lines are arranged discontinuously and alternately.

Images